Data from case-control studies of Alzheimer's disease (AD) were pooled to examine the possible roles of history of depression, anti-depressant treatment and adverse life events as risk factors. History of depression was found to be associated with AD, although the effect was confined to late onset cases. The association held for episodes of depression more than 10 years before AD onset, as well as for episodes occurring within a decade of onset. No association was found with anti-depressant treatment. However, data were only available from two studies, limiting the power of the analysis. Also, no association was found with the three major life events considered in the pooled analysis: death of spouse, death of a child and divorce.